Implications of landslide inventory in susceptibility modeling along a Himalayan highway corridor, India

ABSTRACT Landslide susceptibility is a dynamic phenomenon; it changes with slope intrinsic and transient factors. The study aims to evaluate the changes in the landslides, their size, location, and up to some extent, gradual stabilization impacts on susceptibility assessment. The majority of the models, which are developed to analyze landslide susceptibility, use landslide inventory as a dependent variable to compute the predictors’ significance. Therefore, a slight change in the landslide dataset nullifies the susceptibility zones. Thus, the landslide susceptibility maps become irrelevant to mitigate slope failures. In the present study, we have used two reference period landslide dataset to execute the landslide susceptibility analysis using the frequency ratio method. The model result was validated with a landslide test dataset using the area under the curve (AUC) method. The AUC value is 0.89 and 0.91 for 2013 and 2019, respectively. The study finds that annual updating of landslide inventory is essential for susceptibility modeling for assessment of associated risks and implications of mitigation measures along the highway corridor.